A chiller compressor that rattles its frame, a CRAC unit that hums into the floor slab, a pump skid loosening its own bolts one start-up at a time. These aren't cosmetic problems. In a data center, transmitted vibration shows up later as rack instability, loosened connections, accelerated bearing wear, and in the worst cases mechanical fatigue failures that become an unplanned outage. The fix is rarely "add rubber pads and see." It's a calculation with a right answer, and spring isolators are frequently that answer for the heavier, lower-speed cooling equipment data centers run on.
This guide covers what actually determines whether a vibration isolator works: the relationship between the equipment's operating frequency and the isolator's natural frequency, how to calculate both, and how to specify against real numbers instead of a generic load rating.
Why Isolation Is a Frequency Problem, Not a Material Problem
The instinct is to think in load capacity and material, what holds the weight, what's soft enough to cushion it. That misses the variable that decides whether the isolator works: the relationship between the frequency at which the equipment vibrates and the natural frequency of the system supporting it.
Every mass-and-spring system has a natural frequency at which it wants to oscillate. When the disturbing frequency is near that natural frequency, the system resonates and amplifies vibration rather than isolating it. When the disturbing frequency is well above it, the isolator absorbs the motion and only a small fraction transmits through. The ratio between those two frequencies, not the softness of the material, determines whether an isolator helps, does nothing, or makes things worse. This catches engineers reaching for a "stiffer, more stable-feeling" isolator: a stiffer spring has a higher natural frequency, and if that shift moves it closer to the operating frequency, it transmits more vibration, not less.
The Two Numbers That Determine Whether Isolation Works
Forcing frequency is the frequency at which the equipment actually vibrates, usually tied to rotational speed. A motor at 1,800 RPM produces 30 Hz (1,800 divided by 60); a reciprocating compressor or pump at 600 RPM produces 10 Hz. This comes from nameplate data, not an estimate.
Natural frequency of the isolation system is where the design comes in, and it depends only on static deflection, not on the weight of the equipment:
fₙ ≈ 3.13 / √δ
where fₙ is the natural frequency in Hz and δ is the static deflection (in inches) under the supported load. A heavier unit on a proportionally stiffer spring settles to the same deflection, and the same natural frequency, as a lighter unit on a softer spring. Deflection is the variable that matters, not weight by itself.
| Static Deflection (δ) | Natural Frequency (fₙ) |
|---|---|
| 0.10 in | ~9.9 Hz |
| 0.25 in | ~6.3 Hz |
| 0.50 in | ~4.4 Hz |
| 1.00 in | ~3.1 Hz |
| 2.00 in | ~2.2 Hz |
The pattern is the whole design problem in one line: a lower natural frequency requires more static deflection, a softer spring that compresses further under the same load.
Why the Relationship Between the Frequencies Matters
The ratio of forcing frequency to natural frequency determines transmissibility, the fraction of vibration that gets through the isolator to the structure (or vice versa, for sensitive equipment being protected).
| Frequency Ratio (forcing ÷ natural) | Transmissibility | What It Means |
|---|---|---|
| 1.0 (resonance) | Amplified, can exceed 100% | Isolator makes it worse |
| 1.41 (√2) | 100% | Break-even, no benefit |
| 2.0 | ~33% | Modest isolation |
| 3.0 | ~12.5% | Good, common design target |
| 4.0 | ~6.7% | Strong isolation |
| 5.0 | ~4.2% | Very strong isolation |
Two things follow. A frequency ratio below √2 (about 1.41) doesn't isolate anything and can amplify the vibration; this is a fundamental result of isolation theory, since the system only isolates once the frequency ratio exceeds √2. And the practical target for cooling equipment is 3:1 or higher, where transmissibility drops to roughly 12% and below.
The counterintuitive part: slow equipment is harder to isolate than fast equipment. A 30 Hz forcing frequency only needs a natural frequency around 7.5 to 10 Hz for a 3:1 to 4:1 ratio, a relatively stiff spring with modest deflection. A 10 Hz forcing frequency needs a natural frequency around 2.5 to 3.3 Hz for the same ratio, which demands a much softer spring with far more deflection. As one reference notes, reaching low natural frequencies forces unacceptably high deflections, which is why large, slow-turning cooling equipment is frequently the harder problem.
Worked Example: Sizing an Isolator
Take a chiller compressor section weighing 2,000 lbs on four isolators (500 lbs each), running at 1,800 RPM, a 30 Hz forcing frequency. Targeting a 4:1 ratio (about 6.7% transmissibility):
Required natural frequency: 30 ÷ 4 = 7.5 Hz
Required static deflection: δ = (3.13 ÷ 7.5)² ≈ 0.17 in
Required spring rate: k = 500 ÷ 0.17 ≈ 2,870 lbs/in per isolator
That's a workable, compact spring. 0.17 in of deflection packages easily into a standard isolator housing. Now run the same case on a 600 RPM pump (10 Hz) at the same load and 4:1 target: natural frequency 2.5 Hz, static deflection (3.13 ÷ 2.5)² ≈ 1.57 in, spring rate 500 ÷ 1.57 ≈ 318 lbs/in. A much softer spring, nearly ten times the deflection, a taller isolator with a larger footprint that needs checking for lateral stability, since a tall soft spring can buckle sideways instead of compressing cleanly. The slower machine produced the harder design problem.
Why Steel Springs, and Where They Need Help
They reach low natural frequencies that other materials struggle with, delivering more than an inch of linear, predictable deflection in a compact package where elastomeric mounts can't without becoming impractically large.
Their load-deflection behavior is linear and documented. Rate (k = F/x) is constant and verifiable by load testing, so the natural frequency calculation holds up in service.
They perform consistently across temperature, where rubber-based isolators stiffen in cold and shift their effective natural frequency. This temperature sensitivity of elastomers is well documented: cold environments stiffen elastomeric isolators and push the natural frequency upward.
They last under continuous cyclic load when designed with the same fatigue-life discipline as any continuous-duty spring.
What steel springs don't do well alone is provide damping or block structure-borne noise. With little inherent damping, a system ramping through resonance at start-up or shutdown can briefly amplify; and because the spring is a continuous metal path, it can carry higher-frequency noise straight through. Both are standard problems, solved by pairing the spring with an elastomeric or friction damping pad in a combined assembly, not by abandoning the spring. The damping trade-off is real: more damping protects the system through resonance but slightly reduces high-frequency isolation, which is exactly why a combined assembly is tuned rather than maximally damped.
How to Specify a Vibration Isolation Spring
Identify the disturbing frequency across the full operating range. Variable-speed equipment runs across a band, and the lowest speed is usually the hardest case, so specify the range, not one nameplate number.
Establish the actual load at each mounting point. Total weight divided by number of isolators is a starting estimate; an offset center of gravity means they don't share load equally.
Set the target frequency ratio. 3:1 (~12.5%) is a reasonable baseline; acoustically sensitive spaces may call for 4:1 or higher.
Calculate the required natural frequency and static deflection. This is where forcing frequency, target ratio, and fₙ ≈ 3.13/√δ define the spring's required deflection.
Specify rate, geometry, and stability, and treat noise separately. Check tall soft springs for buckling, and if there's a noise-transmission concern, say so, because it needs a damping element, not a stiffer spring.
Talk to Katy Spring
Katy Spring & Mfg designs and manufactures custom compression springs and spring assemblies for vibration isolation in cooling and mechanical equipment, sized to the equipment's actual operating frequency and per-point load, not a generic catalog capacity. Our engineering team runs the natural frequency and transmissibility calculations with you before tooling starts, and tells you honestly when a spring isolator is the right call and when a combined spring-and-damping assembly is what the application needs.
If you're specifying isolation for cooling equipment, request an engineering consultation or send us your equipment's operating data for a custom isolator quote.
GET A CUSTOM SPRING FORM
Leaders in
custom springs
With all of the ways to communicate (email, text, and fax), we still love phone calls! Call us at 281-391-1888 or contact us online. Follow us on Twitter or Facebook and keep up with the latest on industry news, changes in the raw material market, company news, and other valuable information that will assist you.

Katy Spring
Partners in Precision
Celebrating 25 years of precision
quality + precision for over 25 years.
For over two decades we have delivered quality products made to the specifications of our customers. Our customers’ success is at the forefront, meaning every details matters to Katy Spring.
we serve a wide range of industries & product types
Need a quote? It’s quick and easy, start here:
©2024 All rights reserved. Site by AC